Automated Notion Page Indexing with Vector Embeddings

This n8n workflow automates the process of extracting, processing, and storing content from Notion pages into a vector database for advanced search and retrieval. It triggers on new pages added to a specific Notion database, retrieves the page content, filters out non-text elements like images and videos, and then summarizes the text blocks. The summarized content, along with metadata such as page ID, creation time, and title, is then converted into embeddings using Google’s Gemini API. These embeddings are stored in a Pinecone vector database, enabling efficient semantic search and indexing for knowledge management or AI-powered applications.